JP2008547145A - Apparatus and method for determining a position in a film having film information applied in a time sequence - Google Patents

Abstract

An apparatus is provided for determining a position in film (110) having film information (112, 114) applied in a time sequence, the apparatus storing a reference fingerprint display of film information (112, 114). The reference fingerprint display is formed such that the temporal course of the fingerprint display depends on the temporal course of the film information, and the time scale is associated with the stored reference fingerprint display. And further comprising means (340) for receiving a portion read from the film (110) and means (350) for extracting a test fingerprint display from the read portion. And a test fingerprint display and a reference fingerprint display to determine a position in the film (110) based on the comparison and time scale. And means for comparing (360).
[Selection] Figure 3a

Description

  The present invention relates to an apparatus and method for determining a position in a film having film information applied in a time sequence, for example to synchronize film events and image playback.

  When audio-video data is stored in a fixed format on a data carrier, i.e. film or tape, or on a transmission channel, i.e. wireless or telephone, it will be extended with a new audio format, or other images such as additional services or subtitles Synchronous supplementary services cannot be performed. Thus, for example, with the introduction of a new audio format, it is necessary to generate a new data carrier or film copy with the new audio format.

  FIG. 8 shows an exemplary film 110. Film information such as video information or image 112, each called "frame" or "video frame", and audio information or multiple analog or digital soundtracks 114 are applied in a spatial or temporal sequence during playback A digital soundtrack has an “audio frame” in the digital case. Furthermore, the film 110 has a feed hole 116 used when, for example, reproducing the film.

  Basically, two methods are known as methods for synchronizing supplementary information.

  The first method is to store the time code on a data carrier or another channel connected to the audio signal, using a DTS (Digital Theater System) for movie sound or the like. Examples here include auxiliary data by DAB and mp3. The time code is used for synchronously reproducing sound or additional information from an external data carrier such as a CD using DTS. However, this method has the disadvantage of requiring more space on the data carrier or transmission channel each time a format is added, which may not be available. Films include, for example, tracks for analog sound, Dolby (registered trademark) digital, DTS, and SDDS (Sony (registered trademark) dynamic digital sound). However, because it is a proprietary format, one extension's time code cannot be used for another extension. In the case of expansion, interference with each other is inevitable. As an example, auxiliary data of MP3 is used by various manufacturers for additional information and bandwidth expansion.

  The second method is based on improper use of an analog soundtrack to store time codes, for example, as used in prototype movies with the IOSONO® system. However, the disadvantage of this method is that it is often used as a fallback solution when interfering with other systems because analog tracks exist in all systems, which means that if analog tracks are misused, fallback It means to prevent the possibility of. By automatically switching to the analog track that most movies have, the time code is converted to an analog signal when there is no signal on the "latest" track for Dolby Digital or DTS. It will be played. Therefore, in prototype movies, it is necessary to manually switch between redundant analog playback during pure wavefront synthesis playback as described below, because otherwise time code will be heard through redundant additional speakers. Because it will end up.

  Acoustic wavefront synthesis, or WFS for short, goes beyond the format Dolby®, SDDS or DTS surround approach. WFS attempts to reproduce the vibration of air in the actual situation that makes up the sound throughout the space. Contrary to the conventional method of reproducing over two or more speakers where the mapping of the position of the original sound source is limited to the line between the speakers, wavefront synthesis is a full sound faithful to the original sound source. It is to transmit the field to the space. This means that the virtual sound source can be accurately and spatially arranged and can be surrounded by the sound source because it can be felt as if it were in the space. A system using up to 200 speakers in a movie system and up to 900 speakers in a theater sound system has already been realized.

  Wavefront synthesis is based on Huygens's principle, that is, each point on the wavefront can be considered as the starting point of a fundamental spherical wave. All fundamental waves interfere to generate a new wavefront, which is exactly the same as the original wave.

  Such a sound system was developed at the Fraunhofer Digital Media Technology Laboratory under the name IOSONO (registered trademark) and used in the Ilmenau cinema.

  Therefore, Ilmenau cinema can be said to be an example where wavefront synthesis is performed in two modes.

  In the first mode, the movie is implemented as an “real” wavefront synthesis system. As described above with respect to the second “inappropriate” method, the time code is stored on an analog track of 35 mm film and the WFS sound is played from an external medium such as a hard disk or DVD.

  In the second mode, which is “compatible playback”, the sound stored in each 35 mm film is read and decoded by a Dolby® processor, or using DTS or SDDS. If necessary, the Dolby® processor is automatically switched to an analog track to map multi-channel signals generated via WFS to virtual speakers.

  Since the two modes require different signal paths, it is necessary to split the signal coming from the read head for the analog signal, which leads to further technical effort.

  Thus, in summary, it can be said that there is no room for adding further sync tracks to the current spool of motion picture film, for example for external sound systems or subtitle systems. For all analog and digital movie sound systems available to date, the soundtrack is obtained directly via one or more soundtracks on the film spool, or manufacturer specific time on the film spool. Get by code signal. This means that the two well-known approaches already described require making a new copy of the film, which is usually very expensive. Audio formats such as Dolby® Digital and SDDS allow you to experience the latest audio, but there is still no time code to synchronize with foreign language versions of subtitles or film sound recordings, for example.

  For this reason, Frank Jordan and Jesper Danow's document “Generating Timecode Information from Analog Sound Sources”, 118th Speech Engineering Conference, Conference Papers. 6473, May 28-31, 2005, Barcelona, Spain, proposes generating a time code based on an analog soundtrack. This document describes a system named “Soundtitles” attached to an analog soundtrack of a projector. Based on the edited digital copy of the soundtrack and the analog signal of the film projector, time information or time code is determined by cross-correlation. The system “Soundtitles” is composed of three components. The core module “Sync Tracker” generates a time code signal. The second module “Sync Player” generates subtitles projected using, for example, a beamer. The third module “Clip Player” plays a synchronized audio clip that is communicated to the movie audience via wireless headphones.

  The disadvantage of the prior art described above is that the synchronization and time determination in the film is limited to a search window of, for example, 1 minute, as described in that document. However, particularly in the early stages of film, it is difficult to define or determine the correct window for good synchronization. If the part read or sampled from the film is not in the part of the stored film information used for synchronization, it will remain out of sync or incorrect synchronization will occur. In this case, the movie audience will either not hear the sound on the film or will hear the wrong sound.

Frank Jordan and Jesper Danow's document “Generating Timecode Information from Analog Sound Sources”, 118th Speech Engineering Conference, Conference Paper 6473, 200 28-31 May, Barcelona, Spain

  It is an object of the present invention to provide an efficient concept for determining position within a film.

  This object is achieved by an apparatus for determining a position in a film according to claim 1, a method for determining a position in a film according to claim 20, and a computer program according to claim 21. The

  The present invention is based on the finding that each position of the film generally contains film information specific to this position, and different positions in the film feature extraction include different unique appearances of the characteristics. In other words, different positions of the film contain different “fingerprints”. These fingerprints can in turn be used to determine the position in the film.

  In accordance with the present invention, an apparatus is provided for determining a position in a film having film information applied in a time sequence, the apparatus comprising a reference fingerprint display of film information (FAD, FAD = Fingerabduck dartbelling = fingerprint display). The fingerprint display is formed such that the temporal course of the fingerprint display depends on the temporal course of the film information, and the time scale is associated with the stored reference fingerprint display And means for receiving a portion read from the film; means for extracting a test fingerprint display from the read portion; and a test fingerprint to determine a position in the film based on the comparison and time scale Means for comparing the display with a reference fingerprint display.

  The apparatus and method for determining the position in the film allows determination at any point in the film at any time without having to prepare or change the film itself. The associated time information, time scale, is stored with the stored version of the film. Here, the film is stored in the form of a reference fingerprint display corresponding to feature extraction. Therefore, the required memory space and the period for determining the computing power and / or location can also be reduced.

  The apparatus and method for determining position in film may be used, for example, in an apparatus for generating a control signal for a film event system that synchronizes film events and image playback. Examples of film events include audio sound, subtitles and special effects, which include, for example, airflow, cinema chair shaking, smell, or lighting effects on the side and back walls. May be. Here, with regard to audio results, different audio technologies are possible, such as synchronization of digital surround systems such as wavefront synthesis, along with two different languages such as simultaneously playing back the original version and translation into another language It is. Here, the apparatus or method for determining the position serves to synchronize at the beginning of the film so as to ensure optimal synchronization and / or determination of the position in the film, even under adverse conditions, For example, it provides a high tolerance for jumps during film playback.

  In the above and following examples, when referring to a movie fan or film, the present invention is not limited to movie film viewed by movie fans, but stored on film or other data carriers and memory media, such as magnetic tape or hard drives. Regardless of whether or not it is film information, it generally relates to film or audiovisual signals. Furthermore, the present invention can be used for pure sound systems without video, or can be used for synchronization of pure video material without sound with any event, eg via video ID. .

Preferred embodiments of the present invention will be described below with reference to the accompanying drawings. These drawings include:
FIG. 1 is a basic block diagram illustrating a preferred embodiment of an apparatus for generating a control signal for a film event system,
FIG. 2a is a basic block diagram illustrating an embodiment of an apparatus for performing correlation,
FIG. 2b is a basic block diagram illustrating a preferred embodiment of an apparatus for performing correlation,
Figure 2c. 1 is a diagram showing an exemplary section of a film;
Figure 2c. 2 represents the first variable playback speed and constant test sampling rate in FIG. FIG. 2 shows an exemplary curve of the sound signal of the section of film shown in FIG.
Figure 2c. 3 is the same as that of FIG. 2c. FIG. 2 shows an exemplary curve of the sound signal of the section of film shown in FIG.
Figure 2c. 4 corresponds to FIG. 2c.3 with the third variable playback speed and constant test sampling rate. FIG. 2 shows an exemplary curve of the sound signal of the section of film shown in FIG.
Figure 2d. 1 is a diagram showing two exemplary sections of film,
Figure 2d. 2 is a diagram illustrating an exemplary curve of a reference sound signal of a film,
Figure 2d. 3 is a diagram illustrating an exemplary curve of a test sound signal based on a first playback speed and a constant test sampling rate for a section of film;
Figure 2d. 4 corresponds to FIG. 2 reference sound signal and FIG. FIG. 3 illustrates an exemplary first correlation result from correlation with a test sound signal of 3;
Figure 2d. 5 corresponds to FIG. Figure 2 shows two exemplary sections of a film;
Figure 2d. 6 corresponds to FIG. FIG. 2 shows an exemplary curve of a reference sound signal of film 2;
Figure 2d. 7 is a diagram illustrating an exemplary curve of a test sound signal based on a second playback speed and a constant test sampling rate for a section of film;
Figure 2d. 8 corresponds to FIG. 6 reference sound signal and FIG. FIG. 7 illustrates an exemplary second correlation result from correlation with a test sound signal of 7;
FIG. 3a is a basic block diagram illustrating a preferred embodiment of an apparatus for determining a portion in a film based on a fingerprint display;
Figure 3b. 1 is a diagram showing two sections of the film,
Figure 3b. 2 corresponds to FIG. FIG. 2 shows an exemplary curve of a reference sound signal for two sections of one;
FIG. 4 is a basic block diagram illustrating a preferred embodiment of an apparatus for determining a position in a film based on a coarse determination followed by a fine determination;
FIG. 5a is a basic block diagram illustrating a preferred embodiment of an apparatus for generating control signals for a film event system;
FIG. 5b. 1 is a diagram showing two sections of the film,
FIG. 5b. 2 is a diagram illustrating an exemplary curve of the reference sound signal for the first section of the film;
FIG. 5b. 3 is a diagram showing an exemplary curve of the test sound signal for the second section of the film;
FIG. 5b. 4 corresponds to FIG. 2 reference sound signals and FIG. FIG. 3 shows an exemplary correlation result from correlation with a test sound signal of 3;
FIG. 6a is a basic block diagram illustrating an exemplary film projection system having an apparatus for generating control signals for a film event system and a film event system;
FIG. 6b is a basic block diagram illustrating an example film projection system having an apparatus for generating a control signal having an example audio film event system;
FIG. 7 is a schematic diagram illustrating an exemplary relationship of a time scale to a piece of film information;
FIG. 8 is a schematic diagram illustrating an exemplary film with applied film information.

  In the following description of the invention or the preferred embodiment, the same reference numbers are used for similar or identical components.

  Hereinafter, the present invention will be described in more detail with respect to an embodiment in which a sound signal applied to a film is used as film information. However, the present invention is not limited to this, and this is merely for explanation.

  FIG. 1 shows a basic block diagram of an apparatus for generating control signals for a film event system and the exemplary film 110 described with respect to FIG. 8, wherein the apparatus for generating control signals stores film information. Based on the comparison and time scale; means for receiving 120; means for receiving a section read from the film; means for comparing the read section with the stored film information 112, 114; Means 180 for determining the control signal.

  The stored video information 112, 114 respectively determines, for example, a sound or audio signal, an image signal or a video signal, or a location where an opening is opened, a time when a sound is played or a time when the film is stopped, for example. Contains the display currently seen on the film. The stored audio and / or video signals are for example in a digitized form, preferably in a compressed form that reduces the required memory.

  The advantage of a digitized storage device is simplicity, and in particular, there is no error in the reproducibility of images stored with film information.

  Contrary to the conventional system, as described above, the film is left as it is, and the image in which the film information is stored is generated only once, for example, when the film is created.

  For example, when playing a film via a film playback device such as a film projector, the sound signal contained in the soundtrack 114 is received by the means for receiving 140 and edited by the means for comparing 160, for example, It is sampled at a predetermined sampling rate and sent as a section of a predetermined length or a predetermined number of sampling rates.

  The means 160 is configured to compare the section read from this film with the stored film information, and the means for comparing 160 is configured to compare the read section with all stored information. Although possible, it is preferably configured to compare the read section with the stored section of film information to minimize computational effort. The comparison can be done, for example, via cross-correlation or by calculating the difference, for example by calculating the compressed hash sum and searching it in the database. The comparison can be based on the sound signal alone, on the basis of the video signal alone, on the basis of the comparison of the sound signal and the video signal, or otherwise on the basis of a combination of the above-mentioned feature evaluations. Based on the comparison result and the time scale of the means for comparing 160, the means for determining 180 determines the control signal 190. The film event system is controlled via a control signal 190, and generates, for example, a WFS sound signal or caption based on the control signal 190 in time synchronization with the playback film 110. Thereby, the device for generating the control signal or in particular the means 180 for determining the control signal is an LTC time code format (LTC = horizontal) in which the control signal is standardized according to SMPTE (Movie Television Engineers Association). Time code) or any other standardized time code format.

  Synchronous in time means that the film event system generates, based on the control signal 190, a simultaneous event corresponding to the time on the time scale of the position of the film just played, The time on the time scale is associated with the stored film information.

  Thus, unlike the described embodiment, any film playback device can be used instead of a film projector, and any film format such as silent film (e.g. with synchronization based on video information), analog or A film with a digital soundtrack, a single soundtrack, or several parallel soundtracks can be used, or any other memory medium such as a tape or a hard drive can be used instead of the film Yes, the format cannot be changed or should not be changed, for example compatible with future film playback devices, but other film events need to be synchronized at the same time.

  In the preferred embodiment, the sound signal is used as film information for synchronization. Thus, the section read from the film is sampled at a predetermined sampling rate, hereinafter referred to as a test sampling rate, to generate a test sound signal, and the film information is stored in digital form, and the stored film information is In the following referred to as a reference signal, the test sound signal and the reference sound signal are compared by means 160 for comparing via cross-correlation.

  In one embodiment, the test signal sampling rate and the reference signal sampling rate are unchanged, i.e. constant. The means for comparing 160 may, for example, first generate a first correlation result based on the first test sound signal and the first reference sound signal, determine a first time scale, and then a second Generate a second correlation result based on the test sound signal and the second reference sound signal, for example to determine a time difference or playback speed, or to determine a speed difference compared to a target playback speed or reference playback speed Can be formed to determine a second time scale. Based on this, the means for determining 180 determines, for example, a control signal for synchronizing the film event system.

  However, the disadvantage of constant sampling rate is that the accuracy of determining time or position in the film is further inaccurate and synchronization is reduced because the correlation result decreases with changes in test playback speed. This drawback can be compensated by changing the sampling rate which means the test sampling rate and / or the reference sampling rate.

  FIG. 2 shows a basic block diagram of an apparatus for performing a correlation between a test sound signal that can be played at a variable playback speed and a reference sound signal that is a digitally stored version of the test sound signal. Is provided with means 210 for determining the magnitude of the test playback speed, means 230 for changing the test sampling rate or the reference sampling rate, and means 250 for comparing. The means 230 is configured to change the test sampling rate. Means 230 is generated by sampling test sound signal 270 to generate modified test signal 272 or to change the reference sampling rate to generate a modified reference sound signal based on reference sound signal 274. Formed to change the test sampling rate. Further, the means 230 for changing is such that the deviation between the test playback speed associated with the test sound signal or the reference playback speed associated with the modified reference sound signal 276 is reduced, or the modified test sound signal 272. So that the deviation between the test playback speed associated with the reference sound signal 274 and the reference playback speed associated with the reference sound signal 274 is reduced, or associated with the test playback speed associated with the modified test sound signal 272 and the modified reference sound signal 276. The problem of periodic playback speed or variable playback speed, each of which is configured to change the test sampling rate or the reference sampling rate so that the deviation from the reference playback speed to be reduced, will be described in more detail below.

  Means 250 for comparing the modified sound signal 272 and the reference sound signal 274, or comparing the test sound signal 270 and the modified reference sound signal 276, or comparing the modified test sound signal 272 and the modified reference sound signal 276. , Formed to determine a correlation result 278.

  The embodiment of the apparatus for performing the correlation shown in FIG. 2a can be used as a means for comparing 160 in the apparatus for generating the control signal of the film event system, for example as shown in FIG.

  FIG. 2b shows a basic block diagram of a preferred embodiment of an apparatus for performing a correlation between a test sound signal and a reference sound signal.

  FIG. 2b shows a means 280 for storing a reference sound signal 274 that is a digital version of the test sound signal 270, the reference sound signal 274 previously generated based on a predetermined memory reference playback speed and memory reference sampling rate. ing.

  The test sound signal is played at a variable test playback speed and sampled at a test sampling rate to generate a test sound signal 270.

  The means 210 for determining the magnitude of the test playback speed of the test sound signal 270 controls the means 230 for changing based on the magnitude of the test playback speed. The means 230 for changing controls the reference or sampling rate converter 232 and the variable sampler 234, and the sampling rate converter 232 converts the reference sound signal based on the memory reference playback rate and the memory reference sampling rate to a different memory reference sample rate and The variable sampler 234 is configured to convert to a modified reference sound signal 276 corresponding to a reference sound signal based on a memory reference sampling rate and the variable sampler 234 generates a standard or basic sampling rate to generate a modified test sound signal 272. And configured to sample the test sound signal at a different sampling rate.

  Unlike FIG. 2b, the apparatus for performing the correlation can be configured such that the test sound signal 270 is always supplied to the means 250 for comparison via the variable sampler 234, which then , One of the variable test sampling rates is formed to correspond to a standard or basic sampling rate, and the variable sampler 234 always supplies the means 250 for comparing the reference sound signal 274 via the reference sampling rate converter 232. The reference sampling rate converter 232 is configured to be sent to the means 250 for comparing the reference sound signal 274 so as not to be deformed and to be respectively controlled by means 230 for changing.

  The representation selected in FIG. 2b for supplying the test sound signal 270 to be compared with the modified test sound signal 272 and the reference sound signal to be compared with the modified reference sound signal 276 separately to the means 250 for comparing. These are for explaining another embodiment or another feasibility.

  Thus, for example, in one embodiment where the means for comparing 250 is configured to compare the modified test sound signal 272 and the unmodified reference sound signal 274, the reference sampling rate converter 234 is not required, or The apparatus for performing the correlation of FIG. 2 b does not have a reference sampling rate converter 232. Similarly, the comparing means 250 formed to compare the undeformed test sound signal 270 and the deformed reference sound signal 276 does not have a variable sampler 234.

  In a further embodiment, the means for storing 280 is a means for storing film information, a time scale is associated with the stored film information, and the test sound signal 270 is, for example, film sound. Signal. Next, the apparatus for performing the correlation of FIG. 2b can be used, for example, as the means 160 for comparison of FIG.

Figure 2c. 1 shows a section of an exemplary film 110 having a soundtrack 114 as described above in FIG. Figure 2c. In 1 shows two positions of the film 110, the first position is referred to as a position L _1, called the second position and the position L _2. The two positions L ₁ and L ₂ define a section on the film 110 having a length of ΔL = L ₁ −L ₂ .

Figure 2c. 2 corresponds to FIG. 1 shows an exemplary curve of a test sound signal associated with the section between positions L ₁ and L ₂ shown in FIG. _1, and the time at which film position L ₁ is played is called time T _1, and film position L ₂ is referred to as the time T ₂ time to be played. The time interval ΔT = T ₁ −T ₂ depends on the length of each section and the playback speed v of the film. The following equation ΔT = ΔL / v, or
T ₂ −T ₁ = (L ₂ −L ₁ ) / v
Is true.

  When the test sound signal is sampled at the sampling rate f = 1 / Δt, Δt is a sampling period and ΔT = n · ΔT, and the test sound signal is assumed to be n = 10 as shown in FIG. As exemplarily shown in FIG. 2, it can be represented by a series of n + 1 samples.

When playing a film at playback speed v and sampling rate f = 1 / Δt, the section of film between L ₁ and L ₂ or between T ₁ and T ₂ is divided, for example, by n time intervals, or , N + 1 samples. The following formula is n = ΔL / (Δt · v), or n = ΔL · f / v
Each applies.

  This means that the number of sample periods or the number of samples for a given section ΔL of the film is proportional to the sampling rate f or inversely proportional to the sample period Δt and inversely proportional to the playback speed v. In other words, if n or the number of samples n + 1 is constant, the quotient “f / v” or product “Δt · v” needs to be constant in a section of constant length ΔL.

  In that case, if the initial samples are equal, the individual samples will be equal under the conditions described above.

Correspondingly, when generating stored film information or reference sound signal at memory sampling rate f _memory and memory playback speed v _memory , the stored film information section or test sound signal is, for example, n _memory +1 Represented and stored in reference samples.

To illustrate the fact, FIG. 2 to 2c. 4 shows an example sample or storage of the film section between positions L ₁ and L ₂ at a constant sampling rate f or constant sample period Δt and a variable sample rate, FIG. c2 illustrates exemplary sampling or storage at the first playback speed v ₁ , FIG. 3 shows the sampling or storage of the same section of the film with the _second playback speed v ₂ , FIG. 4 shows the sampling of the same section of the film at the _third sample rate v3. Thus, in this example, v ₁ is, v is half of _2, and, v is twice the _3, a v ₁ = v _2/2 and v ₁ = 2v _3.

Figure 2c. 2 to 2c. All three sound signals shown in 4 have the same sample at position L ₁ or corresponding time T ₁ . Accordingly, correspondingly to FIG. 2 to 2c. As exemplarily shown in FIG. The image information or reference sound signal stored in 2 is represented by n ₁ + 1 = 11 samples, FIG. 3 the same section of the film is represented by n ₂ + 1 = 6 samples, FIG. In 4, the same section of the film is represented by n ₃ + 1 = 21 samples.

Figure 2c. 2 to 2c. As can be seen from FIG. 4, at a constant sampling rate, an increase in playback speed v corresponds to time compression of the sound signal, ie, FIG. When the playback speed v ₁ of 2 is doubled, FIG. 3, T ₂ -T ₁ and n are halved, and the playback speed v is slower with respect to time extension of the sound signal, ie, FIG. When the playback speed v _{1 of 1} is halved, FIG. As shown in FIG. 4, T ₂ −T ₁ and n are doubled.

Figure 2d. 1 and 2d. 2 is simply illustrated in FIG. 1 and FIG. 2c. 2 is supported. Figure 2c. 1 compared to FIG. 1 indicates two additional positions that define a search section or search window for the film and the film information applied to it, where the first position of the search window is denoted L ₀ and the second position of the search window is L ₃ and the section between position L ₀ and position L ₃ is larger than the section defined by positions L ₁ and L ₂ , ie ΔL _window = L ₃ −L ₀ and ΔL = L ₂ −L ₁ ΔL _window > ΔL. Correspondingly, FIG. 2, FIG. In addition to 2, the time T ₀ representing the time associated with the position L ₀ based on the predetermined reproduction speed and the time t ₃ representing the time associated with the position L ₃ based on the predetermined sample reproduction speed are added. Yes.

With respect to the generation of stored film information or reference sound signal and further stored time scale, this defines that time T ₀ defines the time on the time scale associated with position L ₀ , for example, and time T _{0 1} defines the time on the time scale associated with position L ₁ , time T ₂ defines the time on the time scale associated with position L ₂ , and time t ₃ represents the position L on the film. Means that you are defining a time on the time scale associated with ₃ .

  Figure 2d. 3 corresponds to FIG. 2 is supported.

  In the following, FIG. 2 to 2d. With respect to 4, the basic curve of comparison of two signals via correlation or the problem of variable playback speed when comparing two signals is illustrated and described.

This results in FIG. 2 and FIG. 2d. In the optimal case represented in FIG. 3 represents currently read film information applied to the film or test sound signal 270, respectively, FIG. 2 represents the stored film information or the reference sound signal, respectively, and the memory playback speed and the memory sampling rate at which the reference sound signal was generated correspond to the playback speed of the test sound signal and the sampling rate of the test sound signal, respectively. Alternatively, as described above, the quotient of the memory sampling rate f _memory and the memory playback speed v _{memory corresponds to} the quotient of the sampling rate f of the test sound signal and the playback speed v of the test sound signal. In that case, the section of the reference sound signal defined by the reference sound signal or T ₁ and T _2, the test sound signal representative of the section between the T ₁ and T _2, more precisely, their sample sequence, Can be handled accurately, FIG. As exemplarily shown in FIG. 4, a clear maximum or correlation peak can be obtained via correlation.

  The peak position represents the time shift of the test sound signal for each of the reference sound signal or the search window. Based on this, the current time can be determined with respect to the stored time scale.

  Figure 2d. 1 to 2d. In contrast to FIG. 5 to 2d. 8 corresponds to FIG. 7 is the reproduction speed of the test sound signal shown in FIG. An example in which the playback speed of the test sound signal as shown in FIG.

Figure 2d. 5 corresponds to FIG. Corresponds to 1. Figure 2d. 6 corresponds to FIG. This corresponds to FIG. 6 means to represent an exemplary curve of the reference sound signal based on the memory sampling rate f _memory and the memory sample rate v _memory . Figure 2d. 7 corresponds to FIG. 3 or 2d. 6 illustrates an exemplary curve or exemplary sample of a test sound signal based on a test sampling rate f that has not been changed with respect to 6, but based on a modified and slowed playback speed v ′ of the test sound signal.

For the time interval ΔT considered, this is the same as FIG. As shown in FIG. 5, at the same time interval ΔT at the slowed speed v ′, only a short section, ie a section with a short length ΔL ′ due to ΔL ′ = v ′ · ΔT, is reproduced. For a film that has just been regenerated, it means that it reaches the position L ′ ₂ before the position L ₂ . With respect to the reference sound signal and time scale associated therewith, FIG. As shown in FIG. 7, the time scale time T ′ ₂ is associated with the position L ′ ₂ .

  For individual samples of the test sound signal, this means that at low playback speeds, v ′ is the sample period Δt or the corresponding spatial sample, since the “spatial” curve of the test sound signal predetermined by the film soundtrack does not change. Means that each corresponds to a section Δl ′, which is smaller than Δl, so that FIG. 2d compared to FIG. As shown in FIG. 7, the sample of the test sound signal “moves” to the left with respect to the “spatial” signal curve.

If the modified playback speed v ′ is the reverse of the memory playback speed v _memory , the reverse occurs, and at the same time interval ΔT, a long spatial section Δl is played, so the sample of the test sound signal is tested "Move" to the "right" on the signal curve of the "space" curve of the sound signal.

  Therefore, regardless of whether the playback speed is faster or slower than the memory playback speed, the result of the comparison is that the test sound signal and the reference sound signal can be transmitted through two different spatial sections of the film, even under suboptimal conditions. Since it plays, it drops. The result of the comparison becomes worse as the memory playback speed deviates from the test playback speed. When comparing by correlation, the amount of local maxima or peaks is reduced, and since the maximum itself is wider and flatter, the time determination with respect to the time scale becomes increasingly inaccurate until it disappears.

  Under actual conditions, the playback speed of the test sound signal varies not only between, for example, different film projectors, but also in the film. Thus, accurate readjustment is essential to ensure synchronization across the entire film.

  Therefore, in order to represent the same section of film with the same sample, the correlation is performed as described above with the above condition that the quotient of the sampling rate and the playback speed of the test sound signal and the reference sound signal must be the same. The apparatus for changing the sampling rate of the test sound signal or the sampling rate of the reference sound signal in order to minimize the adverse effect of the variable playback speed of the test sound signal.

  In the digital reference sound signal generated at the memory sampling rate, the playback speed is changed by the sampling rate conversion, and the stored reference sound signal 274 generates the reference sound signal at the sampling rate corresponding to the changed playback speed, for example. Are interpolated in the same way.

Figure 2d. 1 to 2d. 8 shows a simplified example, for the sake of clarity, it is assumed that the memory playback speed v _memory corresponds to the normal or common playback speed of the player for generating the test sound signal. However, as described above, the quotient of the sampling rate f and playback speed v must be the same as the reference sound signal and the test sound signal so that the same section of the film can be represented by the same sample as described above. It is the amount that must be. For example, when generating a reference sound signal, a double playback speed can be used when the sampling rate is doubled simultaneously.

  In the embodiment of FIG. 2 b, the means for determining 210 can determine the magnitude of the test playback speed based on the correlation result 278.

  In one approach, for example, to determine whether the deviation between the playback speed of the test sound signal and the reference sound signal is within a predetermined range, the peak speed is compared by comparing the peak amplitude to a predetermined threshold. One correlation result for size determination is used.

  In a preferred embodiment, at least two different reference sound signals based on different reference sampling rates, or at least two different reference sound signals corresponding to different reference playback speeds, respectively, are correlated, for example via quality assessment. Compared to the test sound signal to compare the results, this is based on a known sampling rate and a known memory playback speed, which gives the most similar reference sound signal, and hence the magnitude of the playback speed of the test sound signal, In order to determine from this, it will be described in more detail with reference to FIG. Thereby, different reference sound signals can be formed in succession and compared with the test sound signal, or formed and compared simultaneously.

  In particular, a preferred embodiment of an apparatus for performing correlation generates three reference sound signals based on different reference sampling rates, and the reference sound signal of a medium with three sampling rates is the same as in the previous comparison. The other two reference sound signals are higher than the reference sampling rate or the reference sampling rate of the medium reference sound signal, respectively, based on the reference sampling rate of the reference sound signal that is the best or best matched with the test sound signal It has a low reference sampling rate. This is controlled by means 230 for changing based on the output signal of means 210 for determining the magnitude of the test playback speed. Therefore, the reference sampling rate or the reference playback speed of the reference sound signal can be reliably adapted to the playback speed or the reference sampling rate of the reference test sound signal, respectively.

  FIG. 3a shows a basic block diagram of an exemplary film and apparatus for determining position within the film as shown in FIG.

  The embodiment of the apparatus for determining the position in the film shown in FIG. 3a is, for example, as means 180 for determining the control signal in the apparatus for generating the control signal of the film event system shown in FIG. Can be used.

  The apparatus for determining the position in the film is a memory 320 for storing a reference fingerprint display of film information, wherein the fingerprint display is formed such that the time curve of the fingerprint display depends on the time curve of the film information. A memory 320 associated with a reference fingerprint display in which a time scale is further stored; means 340 for receiving a section read from the film; means 350 for extracting a test fingerprint display from the read section; Means 360 for comparing the test fingerprint display and the reference fingerprint display to determine a position in the film based on the comparison and time scale.

  In a preferred embodiment, the fingerprint display includes a display in the form of spectral flatness, and the time curve of the fingerprint display includes a time curve of spectral flatness.

Figure 3b. 1 shows an exemplary film 110 as shown in FIG. Therefore, when a film is played back at a predetermined playback speed, for example, time scale time T ₁₀₀ is at film position L ₁₀₀ , time scale time T ₁₀₃ is at position L ₁₀₃ , and time scale time T ₁₁₃ is at position L _113. , the time scale of the time T ₁₁₆ is in the position L _116, respectively correspond.

  In one embodiment, in the step of generating a reference fingerprint representation of film information, a fingerprint is determined for each particular space or time portion of the film.

Figure 3b. 2, for example, a first section containing the section of the section or time position L ₁₀₀ ~ position L ₁₁₃ T ₁₀₀ ~ time T _113, the section or time position L ₁₀₃ ~ position L ₁₁₃ T ₁₀₃ ~ time T ₁₁₆ And a second section containing the section. Based on these sections, a fingerprint associated with the section is generated based on, for example, spectral analysis, Fourier transform, or other methods of feature extraction. In a particularly preferred embodiment, the fingerprint includes a spectral flatness γ _x ² calculated from the power density spectral curve, so that the value of spectral flatness is determined for each section and a series of spectral flatness Will depend on the time curve of the film information, for example, the sound signal stored in the memory 320 on an associated time scale.

  The sampling rate, length or duration of the section, or the distance between the next two sections, is determined by demands, for example with respect to the uniqueness or accuracy of the position determination in the film. In general, the longer the section, the clearer the specification of the features, and the higher the sampling rate and / or the shorter the distance between the two sections, the more accurately the position in the film can be determined. The higher the sampling rate, the longer the sections, and the shorter the distance between the sections, the higher the required memory for the reference signal or the computational power signal processing.

  A major advantage of a fingerprint display in the form of spectral flatness is that it requires less memory compared to, for example, complete storage of power density spectra of equal sections. Preferably, a curve or a series of spectral flatnesses are each used as the fingerprint of the section.

  FIG. 4a shows an apparatus for determining a position in a film having film information applied in a time sequence with an exemplary film 110, as shown in FIG.

  The embodiment of the apparatus for determining the position in the film shown in FIG. 4a is a means for determining a control signal, for example in an apparatus for generating a control signal of a film event system as shown in FIG. 180 can be used.

  The apparatus for determining the position is a memory 420 for storing film information applied to the film in a time sequence, wherein the time scale is associated with the stored film information, and a coarse result is obtained. In order to compare a series of samples of the read portion based on a first sampling rate with a first search window of stored film information, and to obtain a fine result indicating the position of the film 110, Synchronization means 460 configured to compare a series of samples of a read section based on a second sampling rate with a second search window of stored film information, The position of the second search window depends on the coarse result, and the first search window is temporally Longer than the search window further comprises a first sampling rate is lower than the second sampling rate, and a synchronization means 460.

  FIG. 5a shows a preferred embodiment of an apparatus for generating a control signal for a film event system with an exemplary film 110 as shown in FIG. 8, wherein the apparatus is an audio signal or test read from the film, respectively. A test sound signal, referred to below as the reference sound signal, to which the time scale is related by comparing the analog sound track applied to the film in the section of the sound signal with the test sound signal and the reference sound signal via the time scale. Based on the stored digital version of the control signal.

  FIG. 5a shows a preferred embodiment of an apparatus for generating a control signal for a film event system having a first film sound sampler 542, the first film sound sampler 542 being a first A / D converter. 544 (A / D = Analog / Digital), the first A / D converter 544 has a correlation between the first feature extractor 552 and the first reference sound signal based on the first sampling rate. A first means 562 for correlating the second means 564 for correlation with a second reference sound signal based on the second sampling rate, and a third reference sound signal based on the third sampling rate. Connected to third means 566 for correlation. The input of the first means 562 for correlation, the input of the second means 564 for correlation, and the input of the third means 566 for correlation are at the output of the sampling rate converter (SRC) 232. Connected.

  The output of the first means 562 for correlation, the output of the second means 564 for correlation, and the output of the third means 566 for correlation are the first means 568 for quality evaluation. Connected to the input. The means 568 for quality evaluation is further connected to a sampling rate converter 232 and a means 570 for sampler selection, and the output of the means 570 for sampler selection is connected to the input of the timer 582. The timer 582 is further connected to a stored soundtrack or means 522 for storing the soundtrack, and the output of the means 522 for storing the soundtrack is connected to the input of the sampling rate converter 232.

  The output of the first feature extractor 552 is connected to the input of a means 554 for comparing features having, for example, a feature classifier and a feature database, and the output of the means 554 for comparing features is Connected to input.

  The output of the timer 582 is connected to or connected to the input of the time code generation means 584 having a time code database, and the output of the means 584 for time code generation is time code smoothing. The means 586 for time code smoothing is connected to the input of the means 586 for generating the time code 592 and the output of the means 586 for time code smoothing is the word clock Connected to the input of a word clock generator 588 configured to output signal 594.

  Optionally, the apparatus for generating the control signal of the film event system further comprises a second film sound sampler 542 ′ connected to the second A / D converter 544 ′, and the second A / D The converter 544 ′ includes a second feature extractor 552 ′, a fourth means 562 ′ for correlation with a fourth reference sound signal based on the first sampling rate, and a second means based on the second sampling rate. The fifth means 564 ′ for correlation with the fifth reference sound signal and the sixth means 566 ′ for correlation with the sixth reference sound signal based on the third sampling rate.

  The output of the fourth means 562 ′ for correlation, the output of the fifth means 564 ′ for correlation, and the output of the sixth means 566 ′ for correlation are the second for quality evaluation. Is connected to the input of the sampling rate converter 232, the output of the second means 568 'for quality evaluation is connected to the means 569 for offset compensation, Further, the means 569 for offset compensation is connected to the means 570 for sampler selection.

  Thereby, the first film sound sampler 542, also called the main sampler, is arranged so that the device for generating the control signal has sufficient time for synchronization. Accordingly, the first film sound sampler 542 provides a pre-delayed signal. During synchronization, the correlation window width or the width of the section of the test sound signal is added. Based on the feed hole on the spool of film, the time difference for the advance delay can be adjusted accurately. Three seconds is recommended as the first criterion.

  In the following, the operating mode of the embodiment of the apparatus for generating the control signal of the film event system will be described in more detail based on the test sound signal generated by the first film sound sampler 542 or its signal processing chain. However, the signal processing of the test sound signal generated by the second optional signal processing chain or the second film sound sampler 542 'corresponds to the first one, so that the offset is simply offset. The compensation means 569 will be described in detail.

  The first film sound sampler 542 reads a sound signal from the film soundtrack or samples the sound signal from the film soundtrack and sends this signal to the first A / D converter 544 for the first. The A / D converter 544 is configured to generate a digital audio signal or test sound signal based on the sampling rate of the first film sound sampler 542 and the playback speed of the film that reads the soundtrack or film information.

  Based on the test sound signal 270, one or more features are extracted or a test fingerprint display is formed. For each feature extraction or fingerprint display, for example, spectral flatness is used as the feature or fingerprint. Next, the test fingerprint display is compared with the reference fingerprint display by means 554 for comparing the features or fingerprint display, and as described above, the fingerprint display time curve depends on the time curve of the film information. And the time scale is associated with a reference fingerprint display stored in the means for comparing features 554, the means for comparing 554 comparing the test fingerprint display with the reference fingerprint display and the time scale. To determine a position in the film or to generate a time code signal 554Z.

  Based on the stored reference sound signal 274, the sampling rate converter generates the same signal at a slightly different sampling rate, i.e. a modified reference sound signal for correlation formed in parallel. As a result, the modified reference sound signal includes a case where the original reference sound signal has the same sampling rate. Therefore, in the description of FIG.

  In other words, the sampling rate converter 232 generates three reference sound signals 276 or modified reference sound signals 276, the first reference sound signal being based on the first sampling rate and the first for correlation. The second reference sound signal 276 is based on the second sampling rate and is supplied to the second means 564 for correlation, and the third reference sound signal 276 is Based on the third sampling rate and fed to the third means 566 for correlation. The sampling rate converter 232 provides a slightly stepped signal at different sampling rates to the correlation or means for correlation 562, 564, 566, where the sampling rate always depends on the previously measured maximum peak. Thus, the noise value from the correlation is adjusted. Each correlation receives a modified reference sound signal of this sampling rate, further correlation receives a slightly lower modified reference sound signal that is one step lower, and further correlation receives a higher stepped sampling rate. To do. This ensures that the sampling rate converter can be tuned or synchronized with changes in the speed of the analog sound signal.

Preferably, the means 522 for storing the soundtrack and the sampling rate converter 232 are configured to use a window width of 2 ⁿ in order to calculate a large calculation window with little effort via a fast Fourier transform (FFT). Is done. Four or more correlations can be calculated in parallel to compensate for sudden jumps during the sound check. The correlation window is largely selected to obtain a large correlation peak. To obtain correlation peak detection accuracy in one sample or sample period, oversampling of the input signal or test sound signal can be used.

  The means 522 for storing the soundtrack outputs a reference sound signal whose length of the correlation window depends on the time code signal 582Z supplied by the timer 582, the correlation window being a search window for searching for the test sound signal. .

  The first means 568 for quality evaluation performs a maximum value search on the cross-correlation result or signal quantity of the signal, and further determines the cross-correlation based on the correlation peak height comparing the cross-correlation with other peaks. It is configured to weight the quality of the correlation results or to determine the quality of individual correlations with respect to peak-to-noise distance.

  Based on the quality assessment, the best quality factor or quality reference sound signal is determined, and based on the position of the peak of the best quality or quality factor reference sound signal, a peak shift with respect to the search window is determined, and For example, it is output as a time code difference between measured and actually effective time codes, or as a relative time code.

  Based on the result of the quality evaluation, the first means 568 for quality evaluation determines the control signal 568A, for example, a sampling rate that distinguishes only three signal values “0”, “+1”, and “−1”. For example, for “0”, the sampling rate of the last sampling rate conversion or correlation is maintained, which means that the correlation result from the modified reference sound signal at the media sampling rate is the highest quality. For “+1”, the sampling rate is increased by one step with respect to the last sampling rate conversion or correlation, since the correlation result from the modified reference sound signal at the highest sampling rate is This is because it has been determined as the highest quality, and “-1” The sampling rate is reduced by one step with respect to the previous sampling rate conversion or correlation, which is the best correlation result or best peak-to-noise correlation between the test sound signal and the modified reference sound signal with the lowest reference sampling rate. It is because it has a distance.

  In other words, based on the best correlation peak being obtained at the (first, second or third) sampling rate, the sampling rate converter is increased or decreased by, for example, the sampling rate delta value. Or it is controlled not to perform sampling rate conversion.

  Thereby, the correlation acts to address two main features. The first is the determination of the position in the film or the time in the film, respectively, based on the time code difference from the correlation, respectively, and the second is the optimal sampling at the optimal reference sampling rate or reference sampling rate, respectively. In order to determine the rate conversion, the size of the playback speed is determined. Here, sampling rate adaptation or generation of adapted sample playback speeds allow for improved correlation results, respectively, thus improving time determination or position determination in the film, and further improving synchronization and prediction .

  The preferred embodiment of FIG. 5 detects signal portions with specific characteristics via signal analysis and suppresses them during synchronization, thus avoiding false detection or synchronization, or time domain Performed to avoid random fluctuations.

  Such features are, for example, signal component loudness or signal “problems”, such as SNR (S / N ratio), PNR (peak to noise), spectral power or power density spectrum, spectral flatness or time sequence. Based on the averaging, signal analysis or problematic component detection can be performed.

  If the peak noise value or the peak noise distance threshold is below, the time code difference can be detected as invalid, for example. Alternatively, if several peaks with similar peak noise distance are determined, the time code difference can be detected as invalid as well.

  Furthermore, since the quality of the correlation of quiet signal components, i.e. small amplitude signal components, is lower than one of the large signal correlations due to high quantization noise during digital sampling, quiet signal components are subject to random fluctuations in the time axis. In order to avoid this, it is suppressed based on a threshold or adaptably. Furthermore, the signal energy becomes an additional quality feature.

  Another example suppresses the problem, because repeated signal components avoid ambiguity, for example, false synchronization.

  Each problematic signal component or portion can be made more prominent as metadata, for example, to suppress these signal components independent of the current correlation quality.

  The means 584 for generating the time code is configured to convert based on the time code signal 582Z of the timer 582, for example, a standardized time code or a standardized time code based on an internal or proprietary time code. Can be converted into a time code signal.

  Timer 582 includes an internal clock (correlation interval or frequency), a coarse audio ID fingerprint or fingerprint display, eg, a time code signal 554Z from a feature determination or fingerprint display, and a means for selecting a determined correlation difference, eg, a sampler. The time code difference signal 570Z determined from the correlation of 570 is controlled. The timer needs to prioritize the correlation signal (highest priority), time code from feature determination and internal clock (lowest priority).

  The means 586 for time code smoothing is useful when the time code signal 584Z is smoothed to avoid, for example, time code that jumps high, or when there is no time code from correlation that compensates for interruptions in analog sound, for example. It is configured to detect an intermediate value. The time code signal 592 generated by the means 586 for time code smoothing is preferably a standardized time code that synchronizes or controls the film event system. However, the time code signal 592 can be used to generate a corresponding sample clock via a slowly adjusting phase lock loop (PLL) if the sound playback system provided is digital. Such phase-locked loops are not subject to patent because they are commercially available as finished devices.

  Optionally, two or more film samplers that provide a time difference offset from the projection lens can be used to improve film robustness or to synchronize inappropriate parts.

  Next, for example, a second film sound sampler 542 'can be used because the second film sound sampler 542' already exists in conventional movie systems. Analog sound interruptions can be linked here by film sound samplers 542, 542 'applied at different locations on the motion picture film, which means that if the film sound interruption is short, the first film sound sampler 542 Alternatively, there is an increased probability that at least one sampler of any of the second film sound samplers 542 'will provide sufficient signals for correlation and associated synchronization.

  Further, optionally, for example, analog sound, Dolby® digital sound (including decoder), DTS digital sound (including DTS decoder), or a combination of these together with different samplers for different sounds, the reference soundtrack and It can be used as a test sound track.

  Here, individual tracks are used for comparison either by averaging, majority voting or prioritization, either automatically or via temporal information metadata generated in the same way as a mono downmix. it can.

  In general, different samplers can be used for different film samplers with different sound formats and / or different time offsets.

  Using a mono downmix has the advantage that when the mono track is used as a stored soundtrack, the stored needs are comparable to storing, for example, 5 channels.

  Some storage devices, i.e. multiple soundtracks, i.e. no downmix, means that all channels are stored independently of each other and then, for example as described above, corresponding comparison or A majority vote needs to be performed to perform the synchronization by using the specific channel, the actual soundtrack and the corresponding channel of the stored soundtrack.

  The initialization phase, the first synchronization and synchronization, is after a sound interruption from two important phases during film projection or during the synchronization of the film event system.

  Thus, since the preferred embodiment initially computes four or more parallel correlations, i.e., does not perform synchronization, this means that four or more reference sound signals of different sampling rates can be used to accurately represent the test sound signal. It means that they are compared or correlated with the test sound signal to determine the sampling rate or sample rate as fast as possible. Here, different sampling rates are tested alternately until one of the correlations gives the best signal noise distance.

  Alternatively or additionally, the first feature extractor 552 and the feature classification means 554 may be used with a database to perform a fine determination of film position or fine time code determination, eg, in a second step by correlation. Provides a coarse absolute time code value that defines a coarse position in the film. As soon as synchronization occurs, the three correlations can be used to synchronize changes in the playback speed of the test sound signal during film projection.

  The accuracy associated with each position in the film or the time associated with the time scale (time code) depends on the sampling rate of the reference sound signal and the sampling rate of the test sound signal, and the higher the sampling rate, the more accurate The position in the film can be determined. However, the lower sampling rate has the advantage that a longer section of the reference or test sound signal can be represented if the number of samples is the same. Accordingly, a preferred embodiment is formed to determine a coarse determination of the position in the film by displaying a long section of the film with a low sampling rate reference sound signal in a first step. The sound signal is obtained by sampling at a low sampling rate. Next, in a second step, based on the coarse position in the film, the high sampling rate reference sound signal and the high sampling rate test sound signal are used for fine position determination in the film.

  In other words, the window length is adapted in performing the correlation. Even if you are oversampling the signal to obtain high time accuracy, a window with a low signal sampling rate is used at the beginning of the search, but the time is roughly detected and followed. A short window is used.

  In the initialization phase, for example, “compatible playback” in “old” audio formats can be performed until the exact position is determined.

  Similarly, “compatible playback” with “old” audio formats can be performed until the correct position is determined once again if synchronization is clearly lost.

  The means 570 for sampler selection and the means 569 for offset compensation are only needed for embodiments using more than one film sound sampler. Thus, for example, the means 570 for sampler selection may be the result or time code difference of the first means 568 for quality evaluation 568Z or the result or time code difference of the second means 568 'for quality evaluation. Determine whether 568Z 'is sent to timer 582 to determine the position in film or time code 582Z. Since the second film sound sampler 542 'samples the test sound signal at different positions on the film, the position where the first film sound sampler 542 samples the film and the second film sound sampler 542' samples the film. The difference between the positions (offset) is compensated by the means 569 for offset compensation, so that the timer 582 stores the time code difference 568Z or the time code difference 568Z ′ last stored, stored in the timer. Regardless of whether the selected time or last stored film position is selected, an accurate time code difference 570Z is obtained.

  Unlike the embodiment shown in FIG. 5a, different reference sound signals with different reference sampling rates can be generated continuously, and further to determine the magnitude of the playback speed of the test sound signal or the optimal reference sampling rate, Can be compared or correlated with test sound signal. Instead, four or more deformed reference sound signals are not only fast synchronized in the early stages, but also during film projection after a large film jump due to film cuts or partial film loss In order to quickly synchronize the current film event system with the current position in the film, it can be compared with test sound signals in parallel or in series.

  Unlike the embodiment shown in FIG. 5, film event system synchronization is performed based on film footage for feature or fingerprint evaluation and test image signal correlation for one or more reference image signals. be able to.

  Thereby, as described above, the correlation of the audio signal and / or video signal can be used to determine the time space of the audio and / or video stream, and the synchronized playback can be controlled by this time determination. it can.

  Instead, the determination of audio and / or video signatures from raw material in the form of audio ID / video ID (ID = identification) coarsens the time of long AV streams to allow synchronization at any location. Can be used to determine.

  The basic approach of the present invention is to further store existing analog sound in digital form for synchronization to motion picture film having an analog sound track via correlation and other characterization. The output signal or control signal of the device for generating the control signal or the synchronization device can be in any time code format, respectively. Preferably, the SMPTE standardized LTC time code format is used. For each motion picture film, at the time of production, a data set needs to be generated for the device for generating the control signal or the synchronization device.

  In production, a separate data carrier is generated for each motion picture film for the above-mentioned means for generating a control signal or a synchronizer. The data carrier includes a digitized analog soundtrack, such as a Dolby® stereo format that can be detected on a spool of film, feature data for the soundtrack, and matching timecode.

  In the following, FIG. 1-5b. An exemplary determination of the time code difference will be described with reference to FIG.

  FIG. 5b. 1 shows an exemplary film 110 having a soundtrack 114, as already described in FIG.

Based on the time code signal 582Z from the timer 582, the reference sound signal 274 is read from the means 522 for storing the soundtrack, and the modified reference sound signal is passed through the sampling rate converter 232 in FIG. Are generated according to 2, which is a film section from the position L ₀ to the position L _3, the time associated with the position L ₀ T _0, or is associated with a position L _3, the corresponding time code and the time T ₃ or timecode Represents.

FIG. 5b. 3 shows an exemplary test sound signal or section of a test sound signal, which is defined by a start time T ₁ and an end time T _{2 and} is generated based on a sampling rate f = 1 / Δt. .

FIG. 5b. 4 corresponds to FIG. 2 modified reference sound signals and FIG. The result of the correlation with the test sound signal section of 3 is shown. FIG. 5b. Time difference ΔT'' = T ₁ -T ₀ between time T ₁ of the start time of the second search window or modified reference sound signal T ₀ and search window or reference sound signal is a time shift, based on this A time code difference or a relative time code is formed. Thus, time T ₁ is the time or time shift of the test sound signal, and the section of the reference sound signal of n = 11 sample length is the test sound signal or reference sound signal of N = 11 sample length and the test. Matches maximally with the correlation of the sound signal and has the maximum as a correlation result.

This eliminates the need for recognition of the absolute time T ₀ or time T ₁ for the quality assessment 568, which means that, for example, the timer 582 knows the last absolute time or absolute time code, respectively, and has further updated the absolute time. Alternatively, only the time code difference 570Z is required to determine the time code. For example, the difference from the peak position with respect to the time to start the search window can be explained. FIG. 5b. 4, the peak is, for example, the first sample, ie, FIG. 3 test sound signals are shown in FIG. The reference sound signal of 2 is shifted by “3 · Δt”, where Δt is a sampling period corresponding to the modified sampling rate.

  Therefore, the time code difference 570Z is composed of the value n = 3, for example. Here, the advantage of the sampling rate or playback speed of the reference sound signal that is respectively adapted to the variable playback speed of the test sound signal is advantageous because Δt is adapted to the playback speed and is related to the search window. A more accurate determination of film position or offset is possible compared to a fixed sampling rate of the reference sound signal, since only multiples of this sampling rate are generated for position determination in the film. is there.

Thus, for example, the time T ₀ of the search window or reference sound signal is equal to the time T ₁ of the previous correlation since the film has been sent forward.

  FIG. 6a shows an embodiment of a film system in which an apparatus 100 for generating a control signal 190 is coupled to a film event system 600, thereby generating a control signal based on the film 110 shown in FIG. The apparatus 100 generates a control signal 190, eg, a time code, that synchronizes the film event system 600.

  FIG. 6b shows a film system having apparatus 100 for generating control signal 190 and wavefront synthesis system 610 as an exemplary film event system, where an embodiment of wavefront synthesis system 610 controls the wavefront synthesis system. Means 620, a digital memory 622 for the wavefront synthesis audio signal, and a plurality of speakers 624 for the wavefront synthesis system. Based on film 110 or analog film soundtrack 114, respectively, means 100 for generating a control signal control signal to enable a wavefront synthesized audio experience using the original analog soundtrack film with a lip sync method. Is generated.

  As an alternative to the wavefront synthesis system 610, it should be understood that other audio systems, such as a digital audio system or a digital surround audio system, can be synchronized via the apparatus 100 for generating the control signal in a Lippy synchro manner.

  FIG. 7 illustrates an exemplary film as shown in FIG. 8, an exemplary digitally stored reference sound signal 720, and time scale association.

  When generating stored film information or a reference sound signal, the analog sound signal is sampled at a predetermined playback speed and a predetermined sampling rate, for example 44.1 kHz, respectively, and a sound part of, for example, 10 milliseconds is so-called audio. Stored as frames, ie, a digital reference sound signal exists as a series of audio frames on the memory. Next, the associated time of the time scale is, for example, in a time code or an audio frame numbered from 0 or 1 in ascending order as the time scale, and the time code TC1 has, for example, an audio frame length of 10 mm Corresponds to audio frame AF1 in FIG. 1 to detect the start time or end time of the audio frame as a time code so that the first audio frame is either 0 milliseconds or 10 milliseconds in the case of seconds is doing.

  The time code typically has a format such as hours: minutes: seconds, and the frames are usually associated with, for example, 24 video frames (movie films) every second.

  Thus, a time scale or time code can associate multiple audio frames with a video frame or define an audio frame as the smallest time scale unit.

  Correspondingly, the time code or time scale can relate four audio frames to one time code, eg referring to TC1 ′ of FIG. 7 including four audio frames AF1 to AF4, or With reference to TC1 of FIG. 7 in which one audio frame AF1 is associated, one audio frame can be associated with one time code. Thus, based on the audio format, the audio frame can also represent a portion of the audio signal overlapping in time.

  The control signal 190 can be formed, for example, as a time code or as a series of pulses, for example, each pulse corresponds to a time scale unit, and the film event system is relative time code to synchronize to the film. Accumulate pulses similar to.

  Further embodiments provide an approach for embedding watermarks in audio and / or video signals, eg, further having an analog sound signal as a fallback, and further simultaneously realizing timecode for another service to be synchronized . The advantage of this approach is that even an “intelligible” audio signal, such as a very quiet sequence or a similar “monotonic” sound, can be fully clock recovered. As a variant of this, basically a full set of appropriate watermark parts is particularly useful in areas where the exact clock rate is searched or the sampling rate is readjusted. However, the decisive drawback of this approach is that it is necessary to change the actual film or create a new version or copy of the film so that the watermark can be embedded in the audio and / or video signal. Is that there is.

  In some situations, the method of the present invention can be implemented in hardware or software. This implementation may be performed on a digital storage medium, in particular a disk or CD, having electronically readable control signals that interact with a programmable computer system so that the method may be performed. it can. In general, the present invention also resides in a computer program product having program code for performing the method of the present invention stored on a machine readable carrier when the computer program product is executed on a computer. Therefore, in other words, the present invention can be realized as a computer program having a program code for executing this method when the computer program is executed on a computer.

FIG. 1 is a basic block diagram illustrating a preferred embodiment of an apparatus for generating control signals for a film event system. FIG. 2a is a basic block diagram illustrating an embodiment of an apparatus for performing correlation. FIG. 2b is a basic block diagram illustrating a preferred embodiment of an apparatus for performing correlation. Figure 2c. 1 is a diagram illustrating an exemplary section of a film, FIG. 2 represents the first variable playback speed and constant test sampling rate in FIG. Figure 2 shows an exemplary curve of the sound signal for the section of film shown in Figure 1; 3 is the same as that of FIG. 2c. Figure 2 shows an exemplary curve of the sound signal for the section of film shown in Figure 1; 4 corresponds to FIG. 2c.3 with the third variable playback speed and constant test sampling rate. FIG. 2 shows an exemplary curve of the sound signal of the section of film shown in FIG. Figure 2d. 1 is a diagram illustrating two exemplary sections of a film, FIG. 2 is a diagram illustrating an exemplary curve of a film reference sound signal, FIG. 3 shows an exemplary curve of a test sound signal based on a first playback speed and a constant test sampling rate for a section of film, FIG. 4 corresponds to FIG. 2 reference sound signal and FIG. FIG. 3 is a diagram illustrating an exemplary first correlation result from a correlation with a test sound signal of 3; Figure 2d. 5 corresponds to FIG. Figure 2d shows two exemplary sections of one film, Figure 2d. 6 corresponds to FIG. Fig. 2d shows an exemplary curve of the reference sound signal of film 2 and Fig. 2d. 7 illustrates an exemplary curve of a test sound signal based on a second playback speed and a constant test sampling rate for a section of film, FIG. 8 corresponds to FIG. 6 reference sound signal and FIG. FIG. 7 is a diagram illustrating an exemplary second correlation result from correlation with a test sound signal of 7. FIG. FIG. 3a is a basic block diagram illustrating a preferred embodiment of an apparatus for determining a portion in a film based on a fingerprint display. Figure 3b. 1 shows two sections of the film, FIG. 2 corresponds to FIG. FIG. 3 shows an exemplary curve of a reference sound signal for two sections of one. FIG. 4 is a basic block diagram illustrating a preferred embodiment of an apparatus for determining a position in a film based on a coarse decision after a coarse position. FIG. 5a is a basic block diagram illustrating a preferred embodiment of an apparatus for generating control signals for a film event system. FIG. 5b. 1 shows two sections of the film, FIG. 2 is a diagram illustrating an exemplary curve of the reference sound signal for the first section of the film, FIG. 3 shows an exemplary curve of the test sound signal for the second section of the film, FIG. 4 corresponds to FIG. 2 reference sound signals and FIG. FIG. 6 is a diagram showing an example correlation result from a correlation with a test sound signal of 3; FIG. 6a is a basic block diagram illustrating an example film projection system having an apparatus for generating control signals for a film event system and a film event system. FIG. 6b is a basic block diagram illustrating an example film projection system having an apparatus for generating a control signal having an example audio film event system. FIG. 7 is a schematic diagram illustrating an exemplary relationship of a time scale to one piece of film information. FIG. 8 is a schematic diagram illustrating an exemplary film with applied film information.

Claims (6)

An apparatus for determining a position in a film (110) having film information (112, 114) applied in a time sequence comprising:
A memory (320) for storing a reference fingerprint display of the film information (112, 114), wherein the time course of the fingerprint display depends on the time course of the film information. A memory (320), wherein a time scale is associated with the stored reference fingerprint display;
Means (340) for receiving a portion read from the film (110);
Means (350) for extracting a test fingerprint display from the read portion;
An apparatus comprising means (360) for comparing the test fingerprint display and the reference fingerprint display to determine the position in the film (110) based on the comparison and the time scale.   The film information is applied to an analog soundtrack on the film, and the means for receiving (340) is configured to receive the analog sound information from the analog soundtrack. The device described in 1.   The means (350) for extracting is configured to calculate a display having spectral flatness as a fingerprint display, wherein the temporal course of the fingerprint display includes a temporal course of the spectral flatness. Apparatus according to claim 1 or claim 2.   4. The method of claim 1, further comprising further means for receiving a portion read from the film, wherein the portion is different from the portion received by the means for receiving (140). The device described. A method for determining a position in a film (110) having film information (112, 114) applied in a time sequence comprising:
Receiving a portion read from the film (110);
Extracting a test fingerprint display from the read portion;
Comparing the test fingerprint display with the reference fingerprint display, wherein the fingerprint display is formed such that a time course of the fingerprint display depends on a time course of the film information (112, 114). And further comprising comparing a time scale associated with the stored reference fingerprint display to determine the position within the film (110) based on the comparison and the time scale. .   A computer program having program code for executing the method of claim 5 when executed on a computer.

Images